Value-Based Contracting in Health Care

Abstract

A growing topic in healthcare in the United States and other countries is the decentralization of risk from the ultimate healthcare payer (insurance companies and government in the United States; national health systems in other countries) to providers of healthcare services. Healthcare providers have traditionally taken clinical risk.1 However, payers are increasingly looking to providers to assume financial risk, in addition to the risk of clinical quality and outcomes of their managed populations. Numerous different types of contracts are being signed between providers and payers: pay for quality; pay for performance; shared risk and shared savings arrangements; bundled payments, accountable care and capitation (full or partial risk). Any contracting entity must decide what is the right form of contract to enter, what contract features to include and what price to offer the payer and what risk the entity is assuming in doing so. The assessment of opportunity, design of the contract terms, pricing, risk management and outcomes evaluation for these contracts are increasingly complex exercises. This chapter covers these issues, including the actuarial mathematics of contract risk assessment and mitigation, taking the reader through the 5 components of a Value-based contract.

1. Introduction

At its most fundamental, health risk (either clinical or financial) is a combination of two factors: amount of loss and probability of occurrence. For the purpose of this chapter we define a loss as having occurred when an individual’s post-occurrence state is less favorable than the pre-occurrence state. Financial Risk is a function of Loss Amount and Probability of Occurrence, or in actuarial terminology, frequency and severity of loss. In the United States health risk has historically been the responsibility of payers (insurers, government programs and employers). Healthcare payers have traditionally managed risk by a combination of pricing, underwriting, and reinsurance, together with claims management. With the enactment of the HMO Act of 1973 (42 U.S.C. § 300e), Managed Care developed in the 1990s as a series of initiatives designed to better manage the health of covered individuals and reduce unnecessary medical claims costs. The original approaches included network management (identifying and contracting with preferred providers who offered either lower fees or lower utilization of services and steering patients to them, either through benefit design or by requiring referrals) and utilization management (pre-authorization or concurrent review of hospital admissions). In a quest for savings these models devolved into restriction of services and denials of care. Because of consumer reaction to the perceived restrictions and denials that resulted from these interventions, managed care plans began to seek other solutions to contain rapidly increasing costs. Techniques that were favored for managing utilization include the implementation of programs that encourage members to take responsibility for their own health, or that aimed to educate physicians in the most cost-effective, evidence-based treatments (Chronic disease management and case management).

The chronic disease management (DM) programs of the early 2000s were implemented by payers and aimed to identify high risk or high need patients, particularly those that were not compliant with their treatments or who had gaps in care. Patient management was usually performed externally, often by telephone, by nurses employed by large disease management organizations. Although attempts were made to involve the patient’s providers, providers were not party to the payer contract. This model reached its peak with a number of Medicare Coordinated Care and Support demonstration programs between 2005 and 2008 [1, 2]. Because of the growth and importance of chronic disease management programs, the Centers for Medicare and Medicaid Services (CMS) of the US Dept. of Health and Human Services (HHS) established a major demonstration project, the Medicare Coordinated Care Project to evaluate 15 different models of care coordination [2, 3]. Although the demonstration program showed some improvement in the quality of care delivered to patients, the lack of demonstrated savings led to a decline in the type of vendor-based disease management programs popular up to that time, and an interest in programs that involved contracting directly with providers to take risk for patient outcomes.

By the end of the first decade of the 21st Century two things began to become clear: first, that these programs were not containing medical trend² and second that the solution to rising costs had to include providers. As a result, CMS’s attention shifted to alternative payment models incorporating providers directly and focusing on a combination of cost, quality and patient satisfaction, an objective expressed by Berwick and others [4] as the “Triple Aim” in a heavily cited article. This shift was a reaction to the quality of care delivered within the US Healthcare system. A 2003 study [5] found that adults in the United States receive the generally accepted standard of preventive, acute, and chronic care only about 55% of the time. Quality of care “varied substantially according to the particular medical condition, ranging from 78.7 percent of recommended care to 10.5 percent of recommended care for alcohol dependence.” Pay for quality was intended increase the frequency of these measures by rewarding physicians for their achievement of evidence-based quality measures (such as screenings, tests for patient populations or adherence to prescriptions). The theory was that closing gaps in care and identifying health issues earlier would lead to reduced utilization of more expensive healthcare services later. The achievement of reduced cost of care in exchange for incentive payments made this a value-based initiative.

Following the failure of the disease management model to demonstrate financial success, Congress has passed a number of laws promoting different value-based initiatives, in addition to initiatives introduced by the Center for Innovation at CMS:

• Medicare Improvements for Patients and Providers Act (MIPPA) 2008;

• Affordable Care Act (ACA) 2010;

• Bundled Payments for Care Improvement (BPCI and its successors) 2011;

• Protecting Access to Medicare Act (PAMA) 2014;

• The Medicare Access and CHIP Reauthorization Act (MACRA) 2015;

• Medicare’s direct contracting model: Global and Professional Direct Contracting Model (GPDC) 2020.

In addition, CMS has introduced a number of alternative payment models (APMs). In these models, providers agree to accept a portion of their reimbursement, often in the form of a share of savings, based on achievement of certain goals, including improved quality, reduced utilization and reduced cost. APMs include Accountable Care Organizations (ACOs) as well as models aimed at specific conditions or provider organizations: Bundled Payments for Care Improvement (BPCI), Comprehensive Care for Joint Replacement, Comprehensive Primary Care, Comprehensive End-stage Renal Disease model, Kidney Care Choices model, and the Oncology Care Model (OCM). CMS’s stated objective is to move the entire health care market toward paying providers based on the quality, rather than the quantity of care they give patients.³

The Health Care Payment Learning and Action Network (HCP-LAN) is a group of public and private health care leaders launched by the U.S. Department of Health and Human Services (through CMS) in March 2015. HCP-LAN aligns public and private sector stakeholders in shifting away from the current fee-for-service, volume-based payment system to one that pays for high-quality care and improved health. HCP-LAN has published estimates of value-based contract penetration in different payer segments. Figure 1 illustrates a study published in 2019 predicting that as much as 100% of care will be delivered via a value-based contract by 2025.

Figure 1.

Estimates of value-based contract growth in different payer segments.

The HCP-LAN 2020 survey of payers indicated that 40.9% of U.S. health care payments, representing approximately 238.8 million Americans and 80.2% of the covered population, flowed through HCP-LAN Categories 3&4 models (shared-risk and population-based payments).

2. Types of value-based contracts

As noted by Werner et al. in a 2021 study [6] “the complexity of the current suite of alternative payment models” and the variety and lack of standardization of different models make value-based contracting challenging. Figure 2 illustrates the development and growth of alternative payment models over time. The following discussion of contract types covers a broad (but not necessarily exhaustive) spectrum: new variations are frequently introduced. Over time, models have become more comprehensive and the risk assumed by providers and healthcare management organizations (HCMs) has increased.

Figure 2.

Risk and VBC contract types. *BPCI: Bundled Payment for Care Improvement; **OCM: Oncology Care Model; ***MSSP: Medicare Shared Savings Program.

Figure 2 illustrates the two dimensions of risk that are accepted by a provider or HCM: the x-axis indicates increasing degrees of financial risk, from none (pay for performance or pay for quality which represent supplemental payments on top of regular provider reimbursement) to capitation (which represents the potential for significant gain but also losses). The y-axis illustrates the extent of the services at risk incorporated in the contract, which may range from a risk limited to a single episode of care only (for example knee surgery) to population risk. Population risk in turn may be limited to certain services only (for example for maternity services those associated with the pregnancy only) to “total cost of care” in which the provider or HCM accepts financial risk for all expenses incurred by the target population.

As we discussed above, the original reimbursement model was fee-for-service: each time the patient received a service from a physician, hospital or pharmacist a bill was generated and then paid by the patient or the payer (or both). As this system began to impose a financial strain on payers, different models evolved, beginning with payment for quality. Payment for quality models addressed the “gaps in care” issue identified in [5], as well as attempting to limit the provision of excess and ultimately redundant services. While these models resulted in improvement in quality metrics (such as HEDIS https://www.ncqa.org/hedis/) they did not lead to significant reduction in healthcare costs. Closely allied to pay for quality models is pay for performance in which physicians are rewarded for patient metrics (such as mammograms for women, eye and foot exams for people with diabetes, etc.).

The big breakthrough in terms of financial risk transfer occurred with disease management programs in the early 2000s. Insurers that purchased disease management programs from vendors needed assurance that the programs would reduce medical cost. Lacking convincing randomized studies, vendors and payers contracted around a financial outcome; initially vendors put a portion of their fees at risk of a favorable financial outcome. Later models allowed vendors to share in actual savings generated (gain-sharing), to the extent that the vendor reduced costs below a target. There are different variations of gain-sharing models, with some being one-sided (only positive savings are shared) while others are two-sided (if costs increase relative to the target, the vendor must reimburse some portion of the excess). More discussion of these models and methods for measuring financial outcomes may be found in Duncan [7].

CMS introduced another value-based arrangement with its Bundled Payment initiative in which organizations entered into payment arrangements that included financial and performance accountability for episodes of care. These models aimed to increase quality and care coordination at a lower cost to CMS. Providers continue to bill CMS in the usual way, with a retrospective reconciliation of claims against a previously agreed upon target price. Depending on which of four payment models the provider enters into, the provider receives a payment that covers hospital only or hospital plus physician services. To the extent that the provider is able to manage the financial risk, it keeps the financial margin (in some models the provider is responsible for reimbursing CMS if costs exceeded target prices). See [8] for a description of the different BPCI models and the results of evaluations.

The Affordable Care Act (2010) [9] introduced Accountable Care Organizations (ACOs): provider groups that accept payment risk for their attributed populations in return for the opportunity to share savings when costs are reduced below an adjusted benchmark. In the original model providers only accepted upside risk (shared savings only). In later models providers could achieve a greater share of savings but at the cost of having to share also in losses. More detail may be found in [10]. ACO arrangements exist among all payers and payer types, including commercial insurers, traditional Medicare and Medicaid. CMS’s Oncology Care Model is a similar initiative but limited to cancer patients undergoing treatment by oncologists.

All these models involve some sharing of risk between the payer and providers. Full risk transfer is achieved with capitated models. With capitation the provider accepts full financial responsibility for all costs of a population (or sub-population, for example primary care only).

3. Five steps to value-based contracting

Value-based contracting requires a clinical organization that is different to the traditional practice management. Several texts discuss necessary re-organization of clinical practice and the necessary infrastructure [11, 12, 13, 14, 15, 16] etc. For the purposes of this chapter we assume that clinical delivery has been optimized and the provider of clinical services is ready to begin the financial modeling required to negotiate contract with a payer.

We illustrate the contract modeling and implementation steps in Figure 2.

Successful value-based contracting requires sophisticated analytics, and at the heart of the analysis is a robust data warehouse that integrates claims data, preferably with clinical data. The importance of claims data is often overlooked by providers, with their focus on clinical data, charts and electronic medical records. Healthcare claims in the US system are the basis of reimbursement, containing valuable information about the nature and diagnosis of a patient’s condition, the treatment applied by the physician or health system, the place of service and (in the case of drugs) the therapeutic class and dosage of a drug. Complete medical and drug claims—claims that include all providers utilized by a population—are essential for financial contracting but are seldom present in provider records: they must be obtained from a payer. Providers rarely have as complete a view of the patient’s care that the payer has (due to its contracts with multiple providers).⁴ Once a robust warehouse has been built, it is possible to begin the five steps to successful value-based contracting (Figure 3).

Figure 3.

Five steps to successful value-based contracting.

3.1 Step 1: opportunity analysis

For any start-up or mature company wishing to enter a value-based contract, the essential first step is to assess the financial opportunity. Payers are subject to multiple new opportunities weekly; a provider or HCM must make a compelling economic case to gain attention. The compelling economic case begins with opportunity. Said differently, does what the provider or HCM intend to contract for address sufficient healthcare spending to be interesting to the payer? Opportunity analysis requires a detailed analysis of healthcare spending on the condition or procedure that the provider or HCM intends to manage. This type of analysis requires detailed healthcare spending (claims) data for the business segment in which the provider or HCM operates. Analysis should address condition prevalence and utilization of the targeted condition(s) and estimate the addressable cost they impose. To gain a payer’s attention the provider/HCM must address an economic concern, which in turn combines two elements:

• Frequency: the condition or procedure must occur with sufficient frequency to be of concern to the payer.

• Severity: the cost imposed by the condition or procedure must be high enough to command the payer’s attention.

Some conditions impose one but not the other of these elements: for example, in an employer population, an episode of stroke is very high cost but occurs with sufficiently low frequency that the average employer may not have experienced a recent stroke in its population. Employees that suffer strokes experience lengthy episodes, during which another payer (such as Social Security disability, or a retirement plan) may become responsible for reimbursement. As a result, the employer may not view strokes as a concern. Cancer, in the other hand, imposes high costs episodically but with cancer diagnoses occurring frequently enough for a payer to be concerned with managing cancer costs.

Modeling opportunity, particularly for individual diagnoses, requires access to large databases. These may be purchased from data vendors, or providers/HCMs may contract with a consultant for this phase of work.

3.2 Step 2: value estimation and economic modeling

Pricing a value-based contract requires an estimate of the value that will be created by a program, device or other intervention (in addition to estimates of the cost of delivery of the VBC solution). Value estimation requires identification of the patient’s current treatment pathway and a projection of an alternative pathway once a VBC solution is implemented. The treatment pathway is a transition or multi-state model that identifies different branches that a patient can follow together with the probability and cost of each different branch. Figure 4 is an example of a simple multi-state model of a specific condition for which the patient can choose to receive treatment in an urgent care setting or a hospital Emergency Department (ED). Depending on the severity of the condition, a patient in the urgent care setting could be sent home or referred to ED. A patient seeking care in the ED could be tested and sent home or, after referral for further evaluation, either sent home or admitted to hospital.

Figure 4.

Current patient pathway.

A detailed claims database will allow the analyst to assess the services, their frequency and the pathway that a typical patient follows. As Figure 4 shows, we associate transition frequencies with the different states, as well as the cost of treatment at different stages. A disruptive device or intervention in this model would reduce the frequency of transition to higher-cost pathways. Figure 4 is a simple pathway; pathways can become extremely complex, in which case some simplification will be necessary. Complexity arises not because of the variety of settings but because the services that the patient receives may be delivered in a different order (for example for some cancer patients, oncology may be delivered first, followed by surgery while for other patients, surgery may be performed first, followed by oncology). Episodes of care that involve physician or auxiliary providers (for example physical therapy) may involve a few treatments over time, to as many as one or two per week.

Once the typical patient pathway is defined and its frequencies and costs have been developed, the analyst can develop an alternative pathway, assuming the provider/HCM intervention has been applied. The alternative pathway illustrates the disruption to the current standard of practice that the provider intervention generates; this may be estimated from prior studies or simply by clinicians who understand the intervention. The difference between the current and proposed pathways, however, is the source of the estimation of the provider’s or HCM’s economic value added. The result of this analysis is an economic model which is the basis of the HCM’s pricing. The economic model is developed by comparing frequencies and unit costs under the current and proposed pathways.

Understanding pathways is a critically important component of the financial estimation process. Providers/HCMs often spend time and effort on the financial estimation phase and assume that the actual work of caring for patients and driving behavior change will take care of itself, if left to clinicians. Clinicians, however, need to know where and how they can perform interventions, with what patients and what outcome to expect. Operationalizing the model to achieve the projected savings is as important as understanding the opportunity. Pathway analysis can provide valuable input to this process because it provides a basis for breaking savings assumptions into drivers/components. We will return below to considering the implementation of a value-based contract.

The Economic Model (Table 1) illustrates the estimation of the value created by the sample intervention illustrated in the pathways in Figure 5, which moves patients from the Emergency Dept. to Urgent Care, as well as more accurately identifies those patients that may safely be sent home after evaluation.

Current patient pathway				Proposed patient pathway
Setting	Patients	Charge	Cost	Patients	Charge	Cost
Urgent care	30	$170	$5100	70	$170	$11,900
Emergency	70	$750	$52,500	30	$750	$22,500
Referred from UC	27	$750	$20,250	35	$750	$26,250
ED evaluation	67.9	$1000	$67,900	32.5	$1000	$32,500
Inpatient transfer	6.79	$30,000	$203,700	6.79	$30,000	$203,700
TOTAL COST			$349,450			$296,850
Intervention			$0		$250	$25,000
Cost/patient			$3495			$3269
Savings %						7.9%

Table 1.

Economic model.

Figure 5.

Proposed patient pathway.

Combining the predicted savings with the cost of delivery of the program allows the Provider/HCM to price its intervention in a manner that allows an appropriate margin for the HCM while also generating an acceptable ROI for the payer. The economic model also allows the HCM to price its contract: in this example the projected savings after intervention charges is 7.9% of projected costs. For a 50/50 gainsharing contract the HCM could each expect savings of 3.95%. This is a point estimate, however, subject to considerable volatility. Before entering into a contract the parties will want to evaluate the uncertainty around the point estimate, which we discuss next.

3.3 Step 3: risk assessment

In Step 2 we created the current and proposed patient pathways, estimated the value created by the HCM and the basic pricing parameters. However, this estimate is a mean; we do not know the variance around the estimated outcome. Variance estimation is important for healthcare models: healthcare claims are highly variable for two reasons. First, the distribution of healthcare claims itself is a convolution of two highly-variable distributions, frequency and severity. Second, outcomes of a healthcare program are subject to performance risk. Step 3 begins with modeling the distribution of the predicted outcome. Additionally, there are multiple variables involved in the predicted outcome; many of these variables can be controlled in order to limit the contract risk. The Risk Assessment step helps the analyst to understand the contribution of individual variables to the predicted outcome and to choose values in such as way as to mitigate some of the inherent stochastic risk of the contracted outcome. Figure 6 shows some of the variables that comprise a value-based contract that an analyst should consider when modeling contract risk.

Figure 6.

Key parameters for a value-based contract.

Figure 6 shows that designing a value-based contract is a complex undertaking. While we will not discuss all the variables in Figure 6, we will discuss some key variables and use them to illustrate the complexity of the modeling that is required as part of the Value-based Contract pricing.

• Attribution: it is important to define precisely those patients for which the HCM or provider will accept risk, and at what point the patient is triggered into the risk group. Attribution can occur on a population basis (for example patients with diabetes) or an episode basis (for example knee surgery). Triggers for these patients generally occur within claims datasets. Occasionally triggers may also be found in electronic medical records (although the lack of integrated medical record/claims data makes modeling difficult in this context). Attribution may also be triggered by the use of a derived marker, for example a grouper model (in the US, Hierarchical Condition Categories (HCCs) or Episode Groupers (for example ETGs)). It is also necessary to use triggers to determine which provider should have accountability for a given patient.

• Acuity: attribution sometimes requires an assessment of patient acuity in cases where the entire population is not managed. Assessment of acuity requires an objective measure such as a predictive model or a grouper model (for example CMS’s HCC Model). A provider/HCM should be wary of clients that want to allow physicians or other clinicians referral or patient self-referral into a program because of their lack of objective evaluation and comparability to a control or comparison population.

• Services: once the patient population is identified it is important to define precisely those services for which the provider/HCM will accept risk. In all cases the question is whether the provider/HCM accepts risk based on claims for a specific condition only, for a subset of services (e.g. PCP capitation), or for “total cost of care?” In each case there will be valid claims included in the risk pool and exclusions. Exclusions are typically those conditions or services that the provider/HCM does not provide or that are managed by a different provider (for example in the case of a diabetes population, a claim for a cancer diagnosis may be excluded because the HCM will not take risk for a non-diabetes related claim).

• Baseline and projection: Many models (the Medicare MSSP ACO model is a good example) rely on comparison of actual outcomes compared with a predicted or projected counter-factual (what would have happened, absent intervention) for the calculation of cost-reduction as the difference between actual and projected costs. A baseline is usually relatively simple to calculate by applying all the contract rules (attribution; services; exclusions etc.) to the payer’s data. As a general rule no contracting party should enter into a risk-based contract without evaluating the population in actual payer data. Estimating what would have happened to the patient population in the absence of intervention is a challenging task, however. This often requires the projection of a cost “trend” or the expected increase of the cost per patient within the treatment population. There are many sources of trend estimates; the MSSP program uses the experience of a non-treated population, adjusted for differences in average risk as its basis for this calculation.

• Stop-loss and truncation: Contracts can be adversely affected by high claim amounts, which occur randomly and unpredictably. For this reason some form of high-cost claim truncation should be considered to limit the contractor’s maximum exposure. Truncation results in amounts in excess of the truncation point defaulting to the payer, which may not be acceptable to a payer. As an alternative the provider/HCM could purchase stop-loss insurance making excess amounts above the truncation point (called the “attachment point in a stop-loss contract) the responsibility of the reinsurer. Stop-loss insurance, particularly for many types of value-based contracts tends to be costly because reinsurers lack experience with many of the very specific types of clinical interventions for which reinsurance is sought, which may result in a provider/HCM deciding to accept the risk of high-cost claims itself. If the provider/HCM has adequate financial resources this may not be a bad strategy, but the provider/HCM should not accept the risk without modeling the potential effect of high-cost claims.

• Risk corridors: an alternative form of risk mitigation is the risk corridor. We discuss this in more detail, with an example below.

Risk assessment requires simulation of the distribution of outcomes. The provider/HCM will contract at a target rate or price assuming its performance will achieve a particular outcome level. In Table 1 this was illustrated as $2,969 per patient. The question to be addressed in the Risk Assessment phase is: what is the confidence interval around this estimate and how may variation be mitigated by choosing different values of the parameters in Figure 6?

Risk mitigation can be illustrated by looking at an example from the Medicare Shared-savings program, assuming that the provider/HCM is considering a contract with both upside and downside risk. The provider will want to maximize its chance of upside gains and minimize the chance of a downside loss (reimburse Medicare). In a recent studies [10, 17] the authors illustrate that even in the absence of an intervention there is a non-trivial risk that a provider will have to reimburse the payer simply because of the stochastic nature of claims, giving rise to the need for Risk Corridors, which are parameters between whose limits no gain or loss is payable.

Figure 7 illustrates this important concept. Note that Figure 7 illustrates stochastic (claims variability) risk only; in addition, the provider/HCM will be at risk of performance variability as well. Figure 7 simulates the outcome (calculated savings assuming no intervention) of 10,000 samples and shows a relatively wide dispersion around the mean. (The mean is zero in this example because we assume no intervention and therefore no savings effect on the population.) With a corridor, the provider is protected against downside risk at the cost of having to give up the opportunity of a gain on the upside. In the example of Figure 7, between 2 and 5% of simulations resulted in losses (reimbursement by the provider/HCM to the payer). The converse is also true: in the majority of cases the imposition of the corridor would have prevented the provider/HCM from receiving a payment despite the HCM having generated savings. If we consider, in addition to the stochastic claims risk, the provider/HCM accepts performance risk as well, the need for sophisticated modeling to understand and mitigate financial risk becomes acute.

Figure 7.

ACO gain/(loss) distribution: 10,000 simulations.

One of the biggest challenges for providers/HCMs entering into value-based contracts is population size. This problem has become especially acute in recent years as providers focus more on specific conditions and sub-populations that may be relatively small or where the condition prevalence results in a small number of target patients. Figure 7 is an example of a 3,000 life population where a target condition could result in only a few hundred patients being managed. The variance in claims of a few hundred patients is significant; the variance may be mitigated with appropriate truncation and risk corridors but in small samples will remain a major risk to the provider/HCM. A number-needed-to-treat analysis could provide some guidance to the contracting parties regarding their potential variance and risk, but the answer is invariably (except in the case of large insurers) that the provider/HCM will need to manage a much larger population than available to be comfortable with the outcomes. In this case the parties should probably consider an alternative contractual form.

The risk corridor is only one variable that can be modeled; modeling the outcomes using the key variables from Figure 6 will give the provider/HCM a better idea of the risk that it undertakes and how to mitigate that risk—for example with risk corridors, different attribution definitions, and stop-loss insurance.

3.4 Step 4: contract terms and operationalizing the model

Once the modeling is completed the contract terms will be known and it should be a straightforward matter to prepare a contract. Once the contract is signed, however, it is important that the provider/HMC prepare an implementation and operational plan with appropriate targets, preferably on a monthly basis. Contractors often lose sight of the fact that they are managing a risk contract, often with a one-year term. If the contractor does not adhere to a plan and falls behind, however, it is often impossible to make up patient engagement and cost-reduction numbers later in the contract year. For this reason a projection of the ultimate results and likely reconciliation on a regular basis is important. For some providers/HCMs (particularly those that are publicly traded) an estimate of the final gain/(loss) will also be required because of the need to set up a balance sheet reserve for any ultimate payable or receivable, and to demonstrate revenue recognition.

Operationalizing the contract also may require sophisticated modeling to identify at-risk patients, alert providers to changes in patient status and report on clinical gaps and gap closure. Delivery of programs that rely on clinical resources is also costly and requires that the contractor maximize efficiency. A workflow system incorporating the latest real-time information for providers (if they are managing patients) or patients (self-management) is essential for efficiency and for achieving contracted outcomes. Monitoring the progress of the contract against the plan and reporting on the key performance indicators identified at Step 2 is essential to achieving successful outcomes.

3.5 Step 5: evaluate outcomes

Some models are relatively simple to administer and reconcile: capitated contracts for example may require no reconciliation because the provider is paid a capitated amount from which the provider derives its margin. Shared savings and bundled payment models, on the other hand, can be complicated to reconcile. One challenge with this type of contract is that reconciliation requires complete data, meaning that run-out claims⁵ are included in the calculation. Allowing for run-out often imposes a delay of 6 months or more post-contract period before complete claims are available. Reconciliation also requires the application of key contract terms: attribution, services, inclusions/exclusions, truncation and corridors etc.

Because value-based contracts are often very different from contract to contract, payers may need to administer contracts manually. This makes final reconciliation difficult both in terms of actual calculation and payments. Reconciliation payments may be delayed as much as 2 years from contract inception. A provider/HCM will need to plan for this delay in receipt of revenue, and have sufficient capital to carry through to the final reconciliation.

4. Payments

Payments are an important part of the Value-based Contract. They represent the result of an intervention, and being part of the operation of the contract, are not a component of the five analytical steps discussed above. Their importance to a contractor and a payer, however, make it important to discuss payments.

A successful contract will result in a payment from the payer to the provider/HCM. Some models such as capitation and bundled payments result in prospective payments: the provider/HCM receives a fixed amount and there is usually no reconciliation or further exchange of funds. For performance-based contracts such as shared-savings or pay-for-performance, a reconciliation will be necessary to calculate amounts owed or owing. Administration of claims for these contracts can be complicated because providers will submit claims in the normal way to the payer, who must then turn off payment (because the provider will be reimbursed from a pool of funds at reconciliation). It is clearly not satisfactory to the provider/HCM to wait 18 months for reimbursement. The challenge of administering partial payments (or payments after the fact) from a typical claims system, particularly in a payer with multiple different contracts, can be challenging to the payer. In many cases these contracts are administered manually. Solutions such as the application of Stochastic Control processes, in which the ultimate settlement payment is continually estimated and payments are made on account of the ultimate payments offer some promise as a way to satisfy provider/HCM need for near real-time payments. That, however, is a topic for a different chapter.

5. Conclusion

Value-based contracts offer providers of healthcare services an opportunity for higher rewards than traditional payment models, but with considerable additional risk. Risk comes in many forms, from definitions to execution. This chapter has not touched on performance risk, which is the province of other professionals, mostly clinical. But aside from clinical risk a provider/HCM that accepts value-based risk is open to numerous other forms of risk. The good news is that with appropriate planning and modeling these risks can be managed and mitigated. Doing so will allow the provider or healthcare management organization to capitalize on a growing trend in healthcare finance.